steffen/self-replicating-neural-networks
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

save weights instead of outputs

Browse Source
This commit is contained in:
Maximilian Zorn
2021-05-17 09:21:01 +02:00
parent c1f58f2675
commit f5ca3d1115
1 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
network.py
View File

@@ -147,10 +147,12 @@ class Net(nn.Module):
# Saving the weights history after a certain amount of steps (aka log_step_size) for research purposes.
# If self-application steps are lower than 10, then append weight history after each SA step.
if SA_steps < 10:
self.s_application_weights_history.append(output.T.detach().numpy())
weights = self.create_target_weights(self.input_weight_matrix())
self.s_application_weights_history.append(weights.T.detach().numpy())
else:
weights = self.create_target_weights(self.input_weight_matrix())
if i % log_step_size == 0:
self.s_application_weights_history.append(output.T.detach().numpy())
self.s_application_weights_history.append(weights.T.detach().numpy())
""" See after how many steps of SA is the output not changing anymore: """
# print(f"Self-app. step {i+1}: {Experiment.changing_rate(output2, output)}")